Duane R. Chartier^a and Fred G. Notehelfer^b

^aConservArt Associates, Inc.; 4823 Berryman Avenue, Culver City, CA 90230-5109

Contrary to popular belief, there is a very great difference between the detection of forgery versus the authentication of works of art. Science is generally very good at producing evidence of falsification but often is equally poor at proving authenticity. The primary reason for these gross differences is that connoisseurship and art history are more strongly involved in the process of authentication than are scientific testing and analysis. There is also a pronounced lack of substantive interaction between art conservation professionals, scientists and art historians.

The case of a recently discovered painting by John Constable will be used to illustrate the difficulties and opportunities of a balanced and systematic approach to the process of authentication. There is much more than science involved in such endeavors and this would not surprise anyone who has attempted to introduce works of art through non-traditional channels.

Great problems arise when the curatorial community is asked to consider works that do not so easily "fit" into a neat art historical period or stylistic pigeonhole. Connoisseurs often will only accept the best works of an artist and discount the inevitable products of the artist's evolution -- less accomplished works. Scientific principles and technical evidence can and must be used in order to elevate the practice of authentication.

Keywords: art authentication, art forgery, fake(s), fakery, art analysis, John Constable

As indicated in the title and abstract of this paper, there is great need for the integration of a scientific approach to the authentication of works of art. What is required is a task that is often paid lip service to but rarely performed on a day to day basis -- true interdisciplinary work. The particular problem is that the special knowledge required is on extreme ends of the academic spectrum -- science and art. The writing of this paper was particularly difficult due to the differences in personal and academic style of the authors. However, that is precisely the synthesis that is required to deal with some of the issues that are presented. The constant interplay of polar approaches is necessary to objectively deal with the "truth" of works of art.

The question of whether a work of art is a fake or authentic may seem relatively simple but there are many complicating factors in the creation, sale, distribution and exhibition of works of art. In fact it is critical to determine the differences between the various gradations of "truth" represented by the continuum⁽¹⁾:

• authentic (by the artist alone)

• authentic (by the artist and others - studio assistants and specialists⁽²⁾ usually not explicitly stated)

• "school of" (by persons in immediate contact and clearly influenced by the artist)

• "circle of" (by persons influenced by the artist)
• "copy of ..." or a "work after ... [a known artist]".
• forgery -- a deliberate attempt to mislead a viewer as to the identity of artist, era, materials, etc.

One must be quite careful to discriminate between works that were initial studies done by the artist in preparation for a more finished or larger scale formal work. These are, of course both authentic, but often vary greatly in style and general treatment.

The gradations of the problem of authenticity can be easily summarized by considering a small case history of a painting owned by a private client that arrived in the ConservArt studios for a conservation treatment. The painting was attributed to Francois H. Drouais (1727-75), a well known French painter. After a thorough examination there were some questions as to why the painting had been lined. The treatment required removal of the lining and when that was done a large black ink stamp on the verso of the painting was found. In German it read: "Copie nach der original #1825 der Kun[??] Gemäldegalerie zu Dresden 1916". An official copy was made in the museum and was accordingly identified as a copy in 1916. Subsequently the painting was lined and then it was misrepresented as being a painting by Drouais. Is this a forgery, a deliberate and fraudulent misrepresentation, or an accidental attribution to the originating artist? Herein lies one of the many problems that are faced by dealers, collectors, and museums. Mis-attribution is not necessarily deliberate and copying is not necessarily forgery. However, the net result in this case was a considerable if not catastrophic devaluation of the work.

The divisions on this scale of veracity can be complex. In the case of prints, connoisseurs have long recognized that original engraved and etched plates have been used over centuries to "pull" new prints. There is an inevitable deterioration of the overall image quality by the progressive mechanical and chemical damages inherent in the printing process.⁽³⁾ These telltale indications of age are also surrogate measures of authenticity and of value. A truly authentic print may rightly be that pulled by the artist or at least, done under their supervision. This is just one example of the potential complexity of authentication questions. Even if one only concentrates on monotypes there are sufficient problems to occupy most art professionals indefinitely.

Forgeries are relatively easy to detect. All one needs is a strongly negative finding to render a judgement as "not authentic". However, the opposite is true for authentication. It is a complex process requiring art historical input, as well as technical/scientific analyses. Can any historic object be deemed 100% authentic? That may be a very thorny philosophic and scientific problem. However, we can do much to refine the process by which we might come to a more balanced approach.

Perhaps the example of the Rembrandt Research Project⁽⁴⁾ might be used to indicate a general direction to pursue for the authentication of a broader range of art and archaeological materials. It is important to consider not only what professional disciplines are required to perform a reliable art authentication but what type of analyses would constitute an acceptable level of proof.

3. ELEMENTS OF AUTHENTICATION

Much of the work of authentication falls into the hands of very few specialists, usually art historians, who have extensively studied the work of one particular artist or group of artists. There are basically two different venues for the work -- the museum and the "outside" world that is composed of dealers, auction houses, collectors, private curators, authenticators, and appraisers. It would require some time to understand and articulate the politics and the operations of these communities and the variations across national borders. That it not the purpose of this paper. However, it is important to note that, regardless of national and cultural differences, there are strong divisions between those within recognized institutions such as museums and those who are not. This is not an insignificant factor in terms of having a work of art accepted as "authentic". Also, there are inherently self-protective and insecure behaviors exhibited by both those on the "inside" and those on the "outside". This is not surprising when one considers the consequences of attribution or rejection of the attribution to a particular artist. It is no understatement that reputations and fortunes hang in the balance, so it should be no surprise that opinions can be vehement and reactions to criticism extreme.

Most "clients" want a positive answer (i.e. a verification that the work is by an important artist) and are loath to be objective about negative findings or "grey areas". The default response of many people who are in a position to authenticate is "NO, THIS IS NOT ..." This is statistically true of most works that are routinely examined by the one of the authors⁽⁵⁾ and such a response guards against foolish mistakes that might ruin one's career or tarnish the reputation of a museum.

Provenance is probably the most critical element in authentication. There is no substitute for an iron-clad provenance back to the hand of the artist touching the canvas. However, this is relatively rare even in the most established of collections. Most works of art that arrive on the art market have changed hands multiple times. Also, wars and political upheavals have drastically changed the owners of record and the paper trail of many works of art. It is safe to say that most works of art fall far short of having impeccable provenance and that is where the interesting problems of authentication begin.

All efforts must be made to track any clues that will improve the provenance of a work in question. When these avenues have been skillfully and completely exhausted then other data may be used to substantiate attribution. Clearly, a keen sense of artistic style and a good knowledge of art history as well as experience is needed to place an unknown work within an era and then, more particularly, within the oeuvre of a particular artist. Having said this, it is important to examine the implicit assumptions in this statement:

• the oeuvre of a particular artist is well enough known to be able to include or exclude particular works
• artists evolve in somewhat programmatic ways, allowing simplifying generalizations of their style
• influences of contemporaries can be assigned and understood
• that artists confine themselves to fairly narrow stylistic limits (at least within certain periods)
• that uncommonly good copies can be easily detected solely by style (this is unlikely).

There may well be many other implicit assumptions that form the essence of what an artistic, aesthetic analysis must become. It seems logical and important to incorporate all other technical information to increase the data available and necessary to make an informed decision.

There are common scientific methods used to examine works of art. Non-destructive techniques such as X-radiography, infrared reflectography, optical microscopy, and ultraviolet fluorescence have been used for decades to examine works of art. There are also many micro sampling techniques that are only minimally destructive and produce analyses of pigments, binding media, fibers and other materials.

Unfortunately, an often overlooked area is the art technology. This is generally the area of expertise of art conservators and some technically minded art historians. A technological examination (an extensive conservation report) will yield potentially very important information such as: canvas type and dimensions, thread count, stretcher type, varnish(es), paint, ground, conservation interventions, rheology (craquelure), etc. These are very useful if put into context but this is only possible, as is the stylistic and art historical analysis, if there is a credible and accessible comparative database.

It is the opinion of the authors that the lack of access or existence of comparative data and/or the lack of creative use of those resources inhibits the development of more reliable and scientifically defensible art authentication. Only by combining approaches can we credibly attack difficult attribution questions with some academic purity. To that end we very briefly present below the findings of an exploration in progress as an example of a collaborative approach that has led to some very exciting hypotheses.

In 1992 a painting appeared in the United States that was sold by an antique dealer as "English Landscape Scene" without artistic attribution. The collector who bought it noticed that it appeared to be a depiction of Willy Lott's house, a favorite John Constable subject. Upon further study it became clear that the painting depicted the subject and composition of Constable's last major painting of the Stour Valley, namely the Valley Farm, his Royal Academy exhibition work of 1835 that now hangs in the Tate Gallery in London. In the Valley Farm Constable depicted his final version of Willy Lott's House, the simple farm cottage pictured in the Hay Wain that stood near Flatford Mill. In this painting Willy Lott's modest home has been turned into a great romantic structure. The house has, in fact, become a brooding, half-timbered building that is fronted by the mill stream. In the foreground at the right a ferryman and a passenger in a boat are moving towards the house. Preceding them are three cows processing through the water towards a small landing to the left of the house where there are several figures as well as another boat with a man leaning into it. Trees flank the house on the right and a pollarded willow on the left. Skimming over the water in the foreground near the ferry is a swallow whose wing has just touched the water and left a small wake. Two moor hens can be seen swimming near the bank at the left front, and a woman appears to be looking out of the door of the cottage towards the ferry and cows. The last of the three cows has her head turned to the right and appears to be lowing.

Constable was extremely proud of this painting, on which he worked even after it was sold to the collector Robert Vernon. He wrote to his namesake, George Constable in 1835, "I have kept my brightness without my spottiness, and I have preserved God Almighty's daylight..."⁽⁶⁾ Unfortunately Constable's high estimation was echoed neither by the critics of the day, nor by scholars of a later generation, who see this painting as a dark and gloomy work revealing that Constable had largely lost his earlier vision of nature and the importance of his country home.

The painting that made its appearance in the United States is almost exactly one-fourth the size of the Valley Farm now in the Tate. The size of the Tate painting is 58" x 49 1/2", the newly discovered painting is 27 3/16" x 23 1/8". It shows a very close conformity to the subject depicted in the Tate painting, but also reveals some important differences. One of the most striking of these lies in the tonality of the work. While Valley Farm is very autumnal, even wintery, in its dark tones, the smaller version of the subject (see the full page Figure 1 that follows) is much closer to Constable's high-summer coloring revealed in the Hay Wain and The Cornfield. The focus of the light on both Willy Lott's house and the clouds in the sky is considerably clearer than in the final painting. The sky has the quality of one of the 1821 sky studies. It suggests having been painted out-of-doors, although the rest of the painting looks more like a studio work. While the tonality of the work differs considerably from the final painting there are other important differences as well. The "Suffolk Girl" painted as the passenger in the ferry seems younger than the figure depicted in the Tate work. The water, itself, is clear and moving. There are also additions and subtractions not in the Tate painting. There is no swallow, whose wings are touching the water; the ash tree in the right middle foreground has none of the veined quality of Constable's late trees, and while the palette knife has been used extensively it has been carefully integrated with the brushwork. Below the woman in the doorway of the house there appears to be a line drawing similar to that of the child depicted in Constable's work Fisherton Mill Salisbury.⁽⁷⁾ Positioned next to the female passenger in the boat, between her apron and basket, there appears to be a soft round bag which is not in the Tate work. While not highly finished in many areas, some parts of the painting show a greater degree of detail than the Tate work. On the lap of the woman passenger there appears to be something that looks like fleece, and in her basket one can see four eggs which are not in the 1835 painting. Among the reeds in the foreground there is a blossoming white convolvulus for which there is no equivalent in the Tate picture, and on the log at the left-front a butterfly which is also missing in the Valley Farm. Finally there is the most important difference of all, the last of the three cows, that closest to the viewer which is lowing in Valley Farm, is clearly depicted as expelling its breath in this painting. Moreover, while the two other cows, those furthest from the viewer, are painted in Constable's manner (usually depicting a large head), the third cow -- particularly the head of this cow -- is painted with a flair and in a manner that is quite different from Constable's treatment of cows elsewhere.

At an early stage the painting was shown to the late Robert Wark of the Huntington who declared it to be a copy of Valley Farm made right in front of the Tate work. Charles Rhyne of Reed College also briefly examined the painting on site and indicated that he too thought it a copy, although he could not identify the hand of the copyist. Both scholars were interested in knowing more about the painting's provenance, which, other than knowing that it had been bought in England shortly before coming to the United States in 1992, remained a mystery. The early judgement by these knowledgeable scholars was therefore that it was a fine copy of the Valley Farm in the Tate Gallery in London.

Still, the painting's quality, particularly the quality of the sky, and the overall effect of the composition remained intriguing. What followed were six years of intensive technical and art historical examination. The technical work on the painting was carried by Dr. Duane R. Chartier of ConservArt Associates in Los Angeles. The art-historical work was undertaken by Professor Fred G. Notehelfer of the Department of History at UCLA.

5. THE ART HISTORICAL EXPLORATION

Art historical examination quickly revealed that while there were a considerable number of sketches associated with the final painting in the Tate, there was no known study that bore the usual close relationship that Constable produced between

(Image Not Available)

Figure 1 - "A Study for Valley Farm" by John Constable ??

Los Angeles, Private Collection

Oil on stretched canvas; 27 3/16" x 23 1/8"

"studies" and final "exhibition works." For example, none of the painted "sketches" included the cows and the ferry with passenger. Malcolm Cormack writing about this in his recent book, Constable, speculated that perhaps in the case of Valley Farm Constable may have "felt his pictures were so much from memory that he needed no trial run."⁽⁸⁾ Still, Valley Farm seems to present a clear anomaly in Constable's normal pattern of operation, for even in the case of his last painting, Arundel Mill and Castle, we have such a study. Why not for Valley Farm?

Trying to deal with the question of this lacuna, led to further discoveries that posed additional matters for consideration. In 1850 the painter Ramsay Richard Reinagle published a letter in The Literary Gazette of London, one of the leading art and literary journals of England, in which he stated: "The late Mr. Constable, R.A., a pupil of mine, exhibited a landscape in the large room of Somerset House [The Royal Academy], in which I painted a group of cattle showing the breath steaming from their mouths; I did them with palette knife to imitate his manner, and he kindly fathered them."⁽⁹⁾ R.B. Beckett, the leading Constable scholar of the past generation, questioned the veracity of Reinagle's statement. He wrote: "As Constable did not use the palette knife for painting until after he had broken off his friendship with Reinagle, the story is difficult to believe."⁽¹⁰⁾ The fact that Reinagle had been forced to resign his position at the Royal Academy two years earlier for exhibiting as his own a work alleged to have been produced by another painter, further discredited Reinagle in the eyes of Constable scholars. It should be noted, however, that when Reinagle's statement appeared in The Literary Gazette in 1850 neither C.R. Leslie his biographer nor Constable's children contradicted Reinagle's statement. Moreover, The Literary Gazette was very cautious about what it published, and the accusations against Reinagle, who was 76 at the time of his forced resignation and had exhibited at the Royal Academy since the age of 13, look rather curious in historical perspective.⁽¹¹⁾ As there are no other paintings in the Constable oeuvre that show cows expelling their breath, this certainly raises an interesting question when a painting of a Constable subject appears that depicts such cattle.⁽¹²⁾ It is even more intriguing when the painting style on the cow expelling its breath seems to differ from the normal style identified with the artist. Curiously, the infrared examination of the drawing under the painted head of the cow shows Constable's usual Suffolk breed structure.

If Reinagle was correct and Constable exhibited a painting at the Royal Academy that depicted cattle expelling their breath, what painting could this have been? Reinagle clearly identifies Constable as an R.A. (Royal Academician) in his letter. Constable was elected to this status early in 1829. It is also worth noting that of all of Constable's paintings exhibited in the "Great Room" of the Royal Academy after his election to full R.A. status, only two have remained unidentified. The first of these is Landscape No. 9 of the 1829 exhibition and the second is Landscape No. 94 of the 1830 exhibition. About Landscape No. 9 of 1829 we know only that he wrote to his friend John Fisher that it was a painting of "a rich cottage."⁽¹³⁾ The Times reviewer reported in his review that it was among the notable smaller paintings.⁽¹⁴⁾ About the 1830 work we know even less. It should be pointed out, however, that the 1829 exhibition was particularly important for Constable as a provisional Royal Academician, and he rather dreaded the "varnishing days," when Academicians touched up their paintings before the exhibition, which he referred to as the "pandemonium" on this occasion.⁽¹⁵⁾ It is also recorded that the sculptor Chantrey decided that he did not like the foreground of Constable's main 1829 painting, Hadleigh Castle, and decided to paint a strong glazing of asphaltum over the foreground to improve it, which Constable spent a good deal of time removing to restore his painting to its earlier state. It is quite possible that Reinagle, whose paintings flanked Hadleigh Castle on both sides, and another of whose paintings hung two spaces from Landscape No. 9 in the Great Room, decided to assist Constable with his cows by showing them to be expelling their breath, and that Constable, who liked the atmospherics of expelled breath, decided to keep the effect on at least one of the cows. Beckett's assertion that Reinagle could not have painted on a Constable painting therefore has to be taken with a grain of salt.

It is also worth noting that Constable regularly exhibited works in Birmingham, and that his pattern for sending works to the Birmingham exhibition was usually to exhibit there a work that had been exhibited in the Royal Academy three years earlier. Often the title of these paintings was changed. It should be noted, therefore, that his exhibition piece in Birmingham in 1832 was titled The Ferry.⁽¹⁶⁾ This work has never been identified either and we can only conjecture that it was a smaller painting due to the room in which it was exhibited. It obviously included a boat with ferryman and passenger. If the "rich cottage" that Constable described to Fisher also included cows and a boat with passenger (ferry), one may be able to link these paintings. This certainly needs to be explored further.

Another intriguing issue involves Constable's friend John Fisher to whom he described the "rich cottage" of 1829. Shortly after the London exhibition, Constable went to visit Fisher and his wife in Salisbury. He stayed for several weeks, and left his daughter and son with the Fishers for some months thereafter (his wife had died previous year). On August 9, 1829, Constable sent the Fishers a painting about which John Fisher wrote: "Your case containing the [rebus that remains undeciphered] arrived just as I was leaving Salisbury. I just got a glympse of its agreeable surface, & came away. I accept of it with the greatest pleasure, as a mark of your friendship, & to my eye, the best specimen of your peculiar art. It is not every man who could have made such a thing a pleasing picture. I intend to hang it over the piano-forte in the first drawing room under the Claude when it comes."⁽¹⁷⁾ In his letter Fisher comments not only on the agreeable surface of the painting, which may suggest that he knew what lay below the surface in earlier versions of the painting, but also on its close affinity to the Claude, which seems to have been the copy of Hagar with Angel that Constable had given him some years earlier. Two weeks later Fisher wrote to Constable, "your great boat looks nobly in the center of the lesser drawing room."⁽¹⁸⁾ While Fisher owned the White Horse, now in the Frick, which also depicted a boat, this painting did not hang in the center of the room but on the wall facing the garden. It would appear then, that Fisher was acquainted with the painting and that it contained a boat. If the painting included both a rich cottage and a boat, we have the essential components for the Valley Farm subject.

In 1838 Fisher's wife Mary wrote to Constable's son who was concerned about the ownership of the painting which was back with Constable at the time of his death. The painting was important to his father who had a "particularly strong attachment to it on account of its being one of his earliest productions and representing a favorite subject." Constable had offered to buy it back for 20, but that she had given it back to the painter instead after her husband's death in 1832.⁽¹⁹⁾ One must remember that in the closing months of 1832 Constable exhibited The Ferry in Birmingham. Since Fisher had died that summer, it may have been an appropriate tribute on Constable's part to exhibit this work which both he and his friend treasured.

Here one has to add some additional evidence for the existence of such a work. C.R. Leslie, Constable's biographer, noted that Valley Farm was painted from an "early sketch."⁽²⁰⁾ Unfortunately he does not indicate what "sketch" this was. It should be noted, however, that on December 9, 1832, eight days after the closing of the Birmingham exhibition, Constable wrote to David Lucas his engraver, "I have sent the long landscape & 2 others. Howard [Henry Howard, Secretary of the Royal Academy] is much pleased with Old Billy Lott's House--so am I."⁽²¹⁾ A few sentences later he wrote, "You may square Old Billy Lott's as soon as you please."⁽²²⁾ R.B. Beckett originally thought that this referred to a mezzotint Lucas was making for Constable.⁽²³⁾ The same interpretation has been given by Leslie Parris of the Tate in a recent publication on Lucas' mezzotints.⁽²⁴⁾ But internal evidence suggests otherwise.⁽²⁵⁾ There is also evidence that Lucas had taken on some of the studio chores (such as squaring) that Constable had earlier assigned to Johnny Dunthorne, his former studio assistant, who died in 1832. Certainly if Valley Farm was squared from another work it would have to look a good deal like the painting we are considering. That the painting from which it was squared was the painting that Constable had exhibited at Birmingham in 1832 as The Ferry also seems logical.

Finally, there is a curious epilogue to the provenance side of this exploration. John Constable died in 1837. His paintings were sold at his "studio sale" at Fosters in 1838. Neither his exhibition piece, Landscape No 9 of 1829, nor the Birmingham Ferry were identified in that sale. Constable's son, John, to whom Mary Fisher had written that the painting about which he inquired after his father's death was as much his as his father's, died in Cambridge in 1841. In January of 1842, Edward FitzGerald, of Rubaiyat fame, bought a Constable at the behest of his portrait painter friend, Samuel Lawrence. This was FitzGerald's second Constable and he wrote to his Quaker poet friend, Bernard Barton in Woodbridge, "it is far inferior in style to the large one, but it is a scene that might please you better, certainly a more Suffolk scene, and it has some things in it as good as can be--the sky especially. It is bout two and a half feet high by two feet wide, and very handsomely framed."⁽²⁶⁾ FitzGerald apologized to Barton that he could not offer the painting to him at the time, but after he brought it to Woodbridge in 1845 it seems to have made its way into Barton's small but impressive collection.⁽²⁷⁾ E.V. Lucas, who described Barton's paintings in his biography of the poet, noted that among his paintings there was "a farm scene by Constable in his earlier and more subdued manner--a very beautiful work."⁽²⁸⁾ When Barton died in 1849 two Constable paintings were in his "sale" in Woodbridge.⁽²⁹⁾ The first was identified in the sale catalogue as Landscape--Farmhouse with Cattle. What painting might this have been? Given Constable's oeuvre as we know it, the only upright oil painting depicting a farm scene that includes a farmhouse and cattle is Valley Farm.

What then is the painting we are examining? Is it a copy of Valley Farm as some Constable scholars have maintained? Or is it the painting of Old Billy Lott's House from which the final work was squared up? Between science and art history we may yet be able to unravel this mystery.

6. THE TECHNICAL EXPLORATION [D. CHARTIER]

PAINTING TECHNOLOGY ASSESSMENT

In an effort to understand the painting technique and to establish the period of the oil painting on canvas, tentatively titled "Study for Valley Farm", a number of non-destructive examinations were carried out. It is important to note that it is rarely, if ever, possible for scientific techniques, alone, to determine the authorship or authenticity of a work. Scientific data can definitively disprove an hypothesis by establishing that materials within an object are later than the attributed date. More commonly, the results are insufficient by themselves to determine authenticity and require the careful integration of historic (provenance) data as well as an intimate knowledge of the particular artist's technique (connoisseurship).

In determining the authenticity of a work, the first technical information of value is an assessment of the characteristics of the material components of the work. This is possible through the use of standard manuals of painting materials and technique and several specific studies of artist's materials.

Canvas and Preparatory Layers

The canvas is a medium weight, plain weave linen with a thread count of 36-38 per inch in both directions. The selvages of the linen are aligned along the vertical and the entire piece of material would be 27 1/2" to 28" in width (depending upon degree of stretching/shrinkage) if laid out flat. This width of canvas is a standard English measure and Constable was known to have used pre-prepared canvas of this width⁽³⁰⁾.

It is not absolutely clear whether the preparatory layer was commercially applied. It is a thin lead white in oil ground. It has not been applied out to the selvage edges (there is about 3/4" of bare canvas on the left hand side and 1/4" on the right hand side). The top and bottom portions of the canvas have ground right up to the cut edge.

The canvas is attached to the stretcher with widely spaced tacks. There are two sets of tack holes in the canvas. The more closely spaced vertical holes up the selvage edges have no tacks at present and indicate that there was a prior attachment to another stretcher or to board. There are no corresponding holes in the present stretcher. The painting is only painted out to the edges on the recto of the stretcher and this indicates that the painting was made for exactly this size. It is possible that the closely spaced holes along the selvages were from tacking out the piece on a board in order to facilitate application of the preparatory layer. This is consistent with colormens' preparation techniques of the period.

There are some indications of a faint brownish or reddish brown secondary preparation that can best be typified as a stain on the original white ground. The traces are seen on all edges but are most distinct on the left hand side and when the untouched areas of the edges of ground are used for comparison.

Stretcher

The dimensions of the stretched painting are 27 3/16" x 23 1/8" (69.1cm x 58.4cm). The stretcher bars are slightly beveled and are 2 1/8" wide x 3/4" thick on the outside edges. The stretcher is a bidirectional pine keyed stretcher with oak keys. The joins at the corner are butted and not mitered at 45.

The canvas, its preparatory layer and the stretcher type used are consistent with Constable's known painting materials and technique.⁽³¹⁾

CHEMICAL ANALYSES

There are relatively few techniques which can be used directly on a painting in order to obtain chemical species data. In general, samples have to be taken and regardless of size this constitutes destructive testing. X-ray fluorescence can be used to identify certain pigments where the X-ray spectrum is particularly strong (mercury and other heavy elements).

The very strong red line running horizontally along the peak of the roof (centered B12.75"/L9") of the left-most ell of the cottage was analyzed and found to be vermillion (a mercury containing pigment). This type of red pigment is consistent with a painting of the 1820's to 1830's and has been noted in many Constable works⁽³²⁾.

Other pigment analyses would require destructive testing and this is not recommended unless there is a very specific and critical issue that could be resolved. In this case it is unlikely that further pigment analyses would produce any surprises.

ULTRAVIOLET FLUORESCENCE

The predominantly green, generalized fluorescence of the painting surface is characteristic of a natural resin varnish. Given the general level of color saturation and the positive solubility test (with ethanol), it is highly likely that the varnish is dammar. The first literature citation indicating dammar being used as a picture varnish is in Lucanus in 1829⁽³³⁾. This implies widespread use prior to that date.

Perhaps the more important information gained from UV fluorescence is the presence of recent overpaints or inpaints. These usually appear as black patches. This is due to two factors:

the relatively new paint medium has not polymerized and cannot absorb significant UV for re-radiation or, more importantly,

there is no varnish over the inpaints/overpaints.

There are only very small areas where the varnish does not fluoresce:

1) some tree trunks and some of the figures at the shore in the middle left hand side immediately to the left of the house (this area seems to have been thinly painted or abraded in a past attempt at varnish removal or from an overly aggressive cleaning).

2) a spot approximately " in diameter centered at B3"/R8.75" which is also slightly visible as an area of retouching in sufficiently raking light and in transmitted light from the verso (see comments for the same area in the infrared discussion - there is a figure of bird contacting the water).

3) the upper body and head of the man poling in the boat in the right foreground (like the area of tree trunks, this area seems to have been thinly painted or abraded in a past attempt at varnish removal or from an overly aggressive cleaning).

All other areas of inpaint cover very minor losses or are isolated on abraded edges. In general, there have been insignificant changes to the painting over time.

INFRA-RED REFLECTOGRAPHY

The complete infrared reflectogram of the painting was assembled from 15 separate images of roughly equal small size. The findings below refer to the detail frames for better resolution. Numbering was done from left to right by row (starting in the top left corner). The images were collected using a Quantex QVC 2500 IR Vidicon camera (this work was performed at the Los Angeles County Museum of art by Alexander Kossolapov who constructed the system⁽³⁴⁾).

The assemblage was captured into a Pentium-based microcomputer using an Imatek® frame grabber and the images were manipulated using Photoshop®. Results were stored on a 250MB Colorado tape backup. The total assembled reflectogram and detail frames 3-15 are not presented here due to publication limitations. In addition there are 4 digitized images of the inscriptions from the verso which are labeled V1-V4 that were used to capture and enhance inscriptions in pencil. The interpretation of these is quite detailed and requires considerable art historical work as well as some creative reading of the script. Analysis of this material is not relevant to the main points of this paper and will be presented in detail in a later publication.

Principal observations are:

the entire area of sky (frames 1, 2, 3) is painted alla prima. There is no underdrawing apparent in the infrared, either in reflected mode or in transmission. This is partially confirmed by the thinness of the paint film in transmitted visible light.

the boles of the trees and major branches (frames 4, 5, 9, 10) are drawn deftly in with little change. The foliage has no indication of underdrawing.

the most apparent of pentimenti is the redrawing of the roof line of the top of the central structure (frames 3 and 4). There is very little other evidence of change or rethinking.

the cows are strongly and decisively underdrawn (frames 7, 8, 9, 12, 13, 14).

frame 14 shows that there was clearly a water fowl of some type which was deliberately overpainted. This is a significant design change which was done at a much more advanced period in the design of the work or may have been done after the work was completed (either by the artist or another party). This frame is inserted below as Figure 2.

Figure 2 - Infrared reflectogram detail of the moor hen on the water to the left of the boat in the lower right foreground (overpainted and not visible in normal light) CLICK HERE for full size picture.

X-RAY

The complete X-ray matrix was assembled from 4 separate images of equal small size. Specific settings used were: 5Ma, 30KV for 2 minutes at a distance of 36". This work was performed at the Los Angeles County Museum of Art on July 9, 1996, by Virginia Rasmussen (Associate Paintings Conservator) and Susanne Friend (ConservArt Associates).

The X-ray data will be discussed in a future paper or monograph dedicated to this investigation. The general findings are covered in the discussion of the infrared work.

OBSERVATIONS / CONCLUSIONS

All materials investigated as well as painting technique are consistent with academic painting practice of the early 19th century.

The sky/clouds in the top left corner are painted very freely, relatively thinly and have no underdrawing. There is evidence of a previous restoration (likely only one campaign given the minimal damages and inpaints). Also, the foreground and midground have been carefully constructed and minimally altered in a manner more akin to an artist's normal working practice and refinement of an original painting in a series of studies rather than a copy.

Some design elements have been changed at the initial stage of underdrawing and at a much later stage (overpainting of the waterfowl in the bottom right corner foreground). Other inpainting is likely associated with the previous restoration(s).

The painting was likely created in the first part of the 19th century and there are technical indications that it may indeed have been painted by John Constable.

The question initially considered by anyone confronted by an unfamiliar work of art is "What is this and who did it?" must be met with the corollary,"What is this not and who else could it be?" Also, in an attempt to examine style and objectify those elements of painting technique that are amenable to such a treatment, it is essential to ask and to try to answer the question, "Is it ... a fake, a copy, a study, or the finished work?" In the case that has been outlined we have come to a point that we can reliably suggest that the painting is not a fake nor is it a copy but rather, it is a study. It may well be a study for Valley Farm but it is probably premature to push the notion that it is a painting by John Constable although there are a great number of concordances. Professor Notehelfer is presently engaged in the demanding task of trying to determine who else the painting may have been executed by and to try to find stronger provenancial links.

John Constable (1776-1837) is arguably the most famous and proficient English landscape painter. Of course, it would be very exciting to participate in the discovery of a fine addition to the oeuvre an important artist. However attractive, we will not fall into the trap of the self-fulfilling prophesy and we will diligently pursue the evidence until some reasonable pronouncement can be academically defended.

7. WHERE DO WE GO FROM HERE?

Much of art authentication seems to be done as a kind of sophisticated cottage industry. Science needs to have a higher profile in art authentication. There is also a real need to examine the process and consequences of authentication so that we may better serve the quest for historical accuracy. There are strong indications that there is a need for an objective, independent agency or committee capable of performing high level authentication research in an atmosphere less dominated by the risks of personal error, institutional prestige, and the blinding effects of money and/or fame.

1. Although there is no single standard, museums tend to use a carefully graded series of attributions from the most secure to the least.

2. In large studios it has been a common practice to have apprentices finish "less important details" such as landscape details in a portrait. A good example of this is the workshop of Peter Paul Rubens.

3. There is a very good and simple discussion in: William Mills Ivins. How Prints Look. Boston: Beacon Press, 1987.

4. The project is unique in that there is a committee of specialists who collectively attribute works. See Grazyna Bastek, "Recenzje: J. Bruyn, B. Haak, S.H. Levie, P.J.J. van Thiel, E. van de Wetering, A Corpus of Rembrandt Paintings. T. I-III." The Hague--Boston--London, 1982-89. (Review of: J. Bruyn, B. Haak, S.H. Levie, P.J.J. van Thiel, E. van de Wetering, A Corpus of Rembrandt Paintings. Vol I-III. The Hague--Boston--London, 1982-89.) Biuletyn historii sztuki Vol. 56 No. 4, pp. 419-428, 1994.

5. During seven years in private practice in art conservation an average of twelve clients per year have come to the ConservArt studios with works of art purporting to be something that they are not. A cursory statistic from this very limited data set is that only about 5% have a reasonable probability of being authentic.

6. R.B. Beckett (ed.), John Constable's Correspondence (Ipswich: Public Record Society, 1968), Vol. V, p. 20.

7. For a reproduction of this watercolor see Graham Reynolds, The Later Paintings and Drawings of John Constable, Plates (New Haven: Yale University Press, 1984), Plate 730 (29.27).

8. Malcolm Cormack, Constable (London and New York: Cambridge University Press, 1986), p. 229.

9. Letter from Ramsay Richard Reinagle to the Editor of The Literary Gazette of London published on May 18, 1850, p. 342.

10. R.B. Beckett, John Constable's Correspondence, Vol. IV, p. 222.

11. The Reinagle case needs further careful examination. Although accused of "plagiarism" by the Academy, he never formally defended himself and questioned the Academy's standing to act as a court in a matter of this kind. He resigned from the Academy when his word was not taken as "a gentleman" to having painted on the painting in question. His two long letters to The Literary Gazette were written to explain the nature of the case. From the points made in these letters it would appear that he could have mounted a successful defense. He noted that the painting in question had been exhibited at Exeter under both his and the other artist, Yarnold's, name, and that having been given the painting by a pupil to show him how to develop the work, he named several witnesses that had seen him paint on the canvas. Later, in London, he was told that there was no such artist, therefore when he sent it to the Royal Academy it went under his name, but he noted that he had not signed the painting as he had the other works submitted. Reinagle, it must be noted, was often critical of the Royal Academy and was regarded as strange and eccentric by many of its members. In 1840 he published a diatribe against the Bank of England that some thought marked him as a radical. The book was titled: "An inquiry into the general system of banking and on the dangerous expedient of a privileged bank to any nation, chartered and founded as the Bank of England." Reinagle therefore had enemies in high places. As the Royal Academy was supported by the King, he was something of an embarrassment to its members. His letter of defense noted that he was not alone in submitting works to the Academy's exhibitions that were not entirely produced by the submitting artist. He named a number of R.A.'s, including the Sculptor Chantrey, who had submitted works that were partially (if not wholly) done by others. It was in this context that he briefly mentioned the Constable painting and its cows. Reinagle had to be very careful about his accusations. To mention something as distinct about a painting as "cows expelling their breath," if no such painting had been exhibited, would have left him open to immediate derision and ridicule.

12. Graham Reynolds has recently published a small sketch, "Landscape with Trees and Cows," which he dates to 1820, that may show two cows expelling their breath. But this is hardly the kind of work that would have been exhibited in the Great Room at the Royal Academy. See Reynolds, The Early Paintings and Drawings of John Constable, Plates (New Haven: Yale University Press, 1996), Plate 1392 (20.98).

13. Constable reported to John Fisher on April 23, 1829, "I have sent the great Castle, such as it is--and a rich cottage." R.B. Beckett, John Constable's Correspondence, VI, p. 244.

14. The Times reviewer noted, "there is a host of small pictures, worthy of observation" among which he included a work by Constable. See Julia Crosby Ivy, Constable and his Critics, 1802-1837 (Woodbridge: The Boydell Press in association with the Suffolk Record Society, 1991), p. 133.

15. See C.R. Leslie, Memoirs of the Life of John Constable (London: John Lemann, 1949), p. 191. Constable had written John Fisher, "I have just got a letter from the Academy --.the Pandemonium opens to the devils themselves on Saturday--in which they [are] allowed every excess for six days (Sunday excepted)." R.B. Beckett, John Constable's Correspondence, VI, p. 244.

16. See Birmingham Society of Arts, Birmingham Society of Arts, Exhibition 1832 (Birmingham: Thomas Knott, 1832), p. 19. The Ferry was No. 171 in the 1832 Exhibition and was hung in the Front Room West. This was one of the lesser rooms and may indicate that this work was not of a substantial size or considered suitable for the Great Room of the exhibition.

17. For Fisher's letter see R.B. Beckett, John Constable's Correspondence, VI, p. 250.

18. R.B. Beckett, John Constable's Correspondence, VI, p. 253. Fisher wrote: "Your great boat looks nobly in the center of the Lesser Drawing Room." There was another reason for Constable to present Fisher with a painting containing a boat. When Constable notified him that he had been elected to full R.A. status in February, Fisher immediately wrote back that he would be there the following day: "Tomorrow I go with you," he wrote, "& call upon your friends. The event is important to me, since my judgement was embarked in the same boat with your success." R.B. Beckett John Constable's Correspondence, VI, pp. 242-43. Constable liked to play with such verbal exchanges and a boat picture would have been an appropriate gift under the circumstances.

19. Mary Fisher's letter can be found in R.B. Beckett, John Constable's Correspondence, VI, pp. 271-72.

20. C.R. Leslie, The Memoirs and the Life of John Constable, p. 257.

21. R.B. Beckett, John Constable's Correspondence, IV, p. 390.

22. Ibid.

23. Beckett writes: "... the squaring of Old Billy Lott's may refer to the cleaning of the margins and the addition of framing lines to prepare it [a mezzotint] for the Appendix." See R.B. Beckett, John Constable's Correspondence, IV, p. 390.

24. See Leslie Parris, John Constable & David Lucas (New York: Salander-O'Reilly Galleries, Inc., 1993), p. 51.

25. A discussion of the internal evidence that questions this use of "squaring" in reference to a mezzotint dealing with Old Billy Lott's House in this passage is being prepared for separate publication.

26. Letter from Edward FitzGerald to Bernard Barton dated London, January 20, 1842. In Alfred McKinley Terhune and Anabelle Burdick Terhune (eds.), The Letters of Edward FitzGerald (Princeton: Princeton University Press, 1980, 4 Vols.), Vol. 1, pp. 300-301.

27. FitzGerald seems to have brought his second Constable to Woodbridge in early 1845. Bernard Barton wrote to his friend John Wodderspoon on January 15, 1845: "Edward has brought home a beautiful Constable sketch, that he gave 20 for, about the size of my head of old Stothard [Barton owned a portrait of the painter John Stothard by Northcote]. I am not sure it is not finer than any finished picture Tom C[hurchyard] or I have..." See letter from Bernard Barton to John Wodderspoon dated January 15, 1845. The British Library. Ms. add 52,524. Quoted by permission of the British Library. Barton's letter indicates that the painting was an upright about the size FitzGerald mentioned in 1842.

28. E.V. Lucas, Bernard Barton and his Friends: A Record of Quiet Lives (London: Edward Hicks, Jr., 1893), pp. 56-57.

29. B. Moulton Catalogue for the Bernard Barton sale, Woodbridge, July 26, 1849, p. 20. British Library, Ms. add 824i49. Quoted by permission from the British Library. Moulton's catalogue of Barton's collection included two Constable paintings. The first was No. 49, Landscape: Farm House and Cattle; the second was No. 76, A Dell at East Bergholt. Unfortunately the size of these works is not indicated in the catalogue. Prof. Notehelfer is preparing a separate paper on Bernard Barton's Constable paintings.

30. See pp. 494-495 of the article by Sarah Cove, "Constable's Oil Painting Materials and Techniques". In Parris, Leslie and Fleming-Williams, Ian. Constable. London: The Tate Gallery, 1991. pp. 493-518.

31. Ibid.

32. Ibid. See Table 1, p. 502.

33. "First Description of Dammar Picture Varnish Translated," Bulletin of the American Group-IIC 7 No. 1, pp.. 8,10, 1966.

34. Alexander J. Kossolapov, "An improved Vidicon TV camera for IR reflectography", in ICOM Committee for Conservation; 10th Triennial Meeting; Washington, DC, USA; 22-27 August 1993; Preprints Volume 1. Paris: ICOM, 1993. pp. 25-31.